Preliminary testing of the prediction that caffeine, a low extraction ration drug (0.1) should be a more sensitive probe of enzyme induction than methacetin, a high extaraction ratio drug (0.9), when excretion of a metabolite (CO2) is measured, has been carried out. A single dose of each was administered to 8 healthy volunteers and 9 epilepsy patients treated with phenytoin, carbamazepine and/or phenobarbital. The 13C carbon dioxide in expired breath was measured by isotope ratio mass spectrometry. The percentages of the dose excreted as CO2 in 2 hr. were compared: 3.22% pus minus 0.86 and 5.54% plus minus 1.59 caffeine was excreted by controls and patients, respectively, compared with 28.6% plus minus 5.8 and 40.0% plus minus 4.2 methacetin. The results in the 2 subject groups were significantly different for both probes (P less than .05). These data do not support the theoretical prediction that the extraction ratio of a drug has a critical effect upon its usefulness in detection of induction of oxidative metabolism via the carbon dioxide breath test. The pharmacokenetics of ethanol have been studied by measurement of blood ethanol concentrations during and after four oral dosing rates and one intravenous infusion to four healthy volunteers. The consumption of breakfast after two hours of oral dosing disrupted the smooth approach to steady state concentrations which cannot be explained as a simple delay in absorption. The data indicate a significant effect of food upon the bioavailability of ethanol probably due to increase in stomach emptying time.